DESCRIPTION (Adapted from Applicant's Abstract: Despite more than a decade of prevention, many people still engage in high-risk sexual behavior. Based on anonymous HIV antibody test site data from San Francisco, the largest group is men who have sex with men and do not use injection drugs. These men are likely to be "high-risk repeat testers", and are accessible through the anonymous testing program. This application will assess a new counseling intervention-applied at the test site-designed specifically to help these men modify their behavior. Modeled after the work of Australian psychologist Ron Gold, the intervention focuses on "self-justifications": specific thoughts, attitudes, or beliefs that "allow" an individual to engage in high-risk activities. The intervention includes a counseling session-aimed at uncovering these justifications-and/or the compilation of a "sex diary", in which individuals identify and record self-justification on each occasion of unprotected intercourse. By bringing these thoughts into awareness, individuals have the opportunity to modify or eliminate them, and in so doing, reduce their frequency of unprotected sex. The UCSF AIDS Health Project in collaboration with the SF Department of Public Health and the UCSF Center for AIDS Prevention Studies, proposes a prospective randomized study of this intervention. The study will comprise four groups of 114 men (N=456) who have tested at least twice before and who admit to unprotected anal insertive or receptive intercourse within the past year. Group 1 will receive usual pre-and post-test counseling, as will all other groups. Group 2 will receive the new intervention, consisting of both the additional counseling session and the assigned "sex-diary" homework. Group 3 will receive the additional counseling session only, and Group 4 will receive the assigned "sex-diary" homework only. Data will be collected at baseline, 6 and 12 months and will include sexual behavior history, alcohol and drug use, presence of STDs, and attitudes towards condom use. This design should achieve the study's four goals: 1) to describe the types and prevalence of risk behaviors in a high-risk population; 2) to describe the cognitive precesses these men employ when they engage in unprotected anal intercourse; 3) to apply a cognitive-behavioral intervention to address specific thought processes at the time of intercourse; and 4) to reduce risk behavior by 25 percent in the intervention groups. If successful, the intervention will be incorporated into an HIV testing protocol for high risk repeat testers.